Hello folks,

I'm fairly new to the Sun hardware side of things, though I have quite
a bit of Linux experience.

I have a box here (Ultra Enterprise 1), which I need to access.  I do
have the root password (but no other usernames/passwords, which is why
I can't SSH in).

My problem is that I am currently unable to get console access to it. 
There is no monitor or keyboard hooked up to the machine.  I was
trying to get access through a serial connection, but I have thusfar
been unable to succeed in that.

My question, I guess, is:  Is it possible to get a console session to
provide me  with a login prompt, without a monitor and keyboard
attached to the Sun box?  And if so, how?

I've tried HyperTerminal, Minicom, cu, and none of them provide any
sort of output from the Sun box.  They say they're connected, but I
see nothing.  I even tried to power-cycle it, thinking that maybe that
would get some screen activity, but to no avail.

I have some other Sun boxes around.  I was thinking of stealing a
keyboard from one of them.  Would the be any problem unplugging a
keyboard from a live system and would it be of any benefit?  or
whatever reason, none of the other Sun boxes have the same type of
monitor connection, so I don't know if I'll ever be able to get that
functional.

I'm out of ideas here.  Thanks in advance for your help.

Jordan

On Tue, 26 Apr 2005, Jordan D. wrote:

> My problem is that I am currently unable to get console access to it.
> There is no monitor or keyboard hooked up to the machine.  I was
> trying to get access through a serial connection, but I have thusfar
> been unable to succeed in that.

Hmm.  That should work.

> My question, I guess, is:  Is it possible to get a console session to
> provide me  with a login prompt, without a monitor and keyboard
> attached to the Sun box?  And if so, how?

Yes, it is possible.  You'll need a null-modem cable, and set your
comms to 9600 8N1.

> I have some other Sun boxes around.  I was thinking of stealing a
> keyboard from one of them.  Would the be any problem unplugging a
> keyboard from a live system and would it be of any benefit?  or

You might drop those Suns to the OK prompt, as keyboards aren't designed
to be hot swappable (then again, you might be lucky...).  It might help,
but when you do, make sure you hold donw Stop-N when booting.  This will
reset the OBP variables to their default.  Disconnect the keybaord and
then try again.

HTH,

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

wraezor@gmail.com (Jordan D.) writes:

>My question, I guess, is:  Is it possible to get a console session to
>provide me  with a login prompt, without a monitor and keyboard
>attached to the Sun box?  And if so, how?

If no keyboard was attached when booting, then the Serial A
port will act as a console (9600,8 bit, 1 stop bit, no parity)

>I've tried HyperTerminal, Minicom, cu, and none of them provide any
>sort of output from the Sun box.  They say they're connected, but I
>see nothing.  I even tried to power-cycle it, thinking that maybe that
>would get some screen activity, but to no avail.

That would be expected, yes, at the proper speed.

Unless the console speed was reset.

>I have some other Sun boxes around.  I was thinking of stealing a
>keyboard from one of them.  Would the be any problem unplugging a
>keyboard from a live system and would it be of any benefit?  or
>whatever reason, none of the other Sun boxes have the same type of
>monitor connection, so I don't know if I'll ever be able to get that
>functional.

Unplugging a keyboard may through the other Sun system into the
boot prom.  First you should run:

	kbd -a disable

on that system; that will disable the keyboard break.

Casper
-- 
Expressed in this posting are my opinions.  They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.

Okay, I'm just a little slow.  :)  It's working now.

What tipped me off was you saying "Serial A" port.  Got me
thinking...hmm..I only remember seeing one port.  Turns out I wasn't
plugging into Serial port A or B.  But instead a port labelled "/ /"
which it turns out to be the parallel port.  Ah well, you learn
something new everyday.

Thanks to both of you for the help.

wraezor@gmail.com wrote:

>Okay, I'm just a little slow.  :)  It's working now.
>
>What tipped me off was you saying "Serial A" port.  Got me
>thinking...hmm..I only remember seeing one port.  Turns out I wasn't
>plugging into Serial port A or B.  But instead a port labelled "/ /"
>which it turns out to be the parallel port.  Ah well, you learn
>something new everyday.
>
>Thanks to both of you for the help.
>
>  
>
Sheesh!!! You're lucky you didn't fry something!  Or maybe you did and 
haven't found it out yet.

On Wed, 27 Apr 2005 17:12:02 -0400, Richard B. Gilbert wrote:

> Sheesh!!! You're lucky you didn't fry something!  Or maybe you did and 
> haven't found it out yet.

They really didn't need that TTL parallel port anyway.

Well, I don't really know if I did wreck anything, I suppose.  We never
have (and likely never will) used the parallel port.

It was more for curiosity sake than anything, that I wanted to get on
there.  Sun is a bit of a mystery to me so I wanted to explore.  The
machine is fairly unimportant in our environment so it wouldn't be a
great loss if it died.

On Thu, 28 Apr 2005 09:02:14 -0700, wraezo wrote:

> Well, I don't really know if I did wreck anything, I suppose.  We never
> have (and likely never will) used the parallel port.

I don't have the schematics for the system board but parallel ports are
generally TTL circuits (0 - 5VDC).  The serial ports using RS-232 swing
their voltages between -12VDC and +12VDC.  Applying -12VDC to the output
of a TTL circuit would probably destroy it.

> It was more for curiosity sake than anything, that I wanted to get on
> there.  Sun is a bit of a mystery to me so I wanted to explore.  The
> machine is fairly unimportant in our environment so it wouldn't be a
> great loss if it died.

The Ultra 1 serial ports are labeled TTY-A and TTY-B but Sun violated
common practice for DTE ports by making them DB-25S connectors rather than
DB-25P.

>>>>> Dave Uhring <daveuhring@yahoo.com> writes:

> The Ultra 1 serial ports are labeled TTY-A and TTY-B but Sun
> violated common practice for DTE ports by making them DB-25S
> connectors rather than DB-25P.

Is this really so? I was under the impression that since the serial
ports (or at least port A) are mainly meant for plugging terminals
into, they are actually configured as DCE-ends. (I never checked the
specs though, and may be mistaken; incidentially however, on Ultra-5
and 10 Models (which are otherwise mainly PeeCees with SPARC CPUs)
port A ist still DB-25S while B is a DE-9P, and the logical conclusion
seems to be that one is configured as DCE-end for OBP-access and the
other as DTE to be compatible with the rest of the PeeCee-world...)

-- 
Stefan Jankowski
Sysadmin, University of Freiburg, CS Dpt.
s/spambucket//
s/dave\.null/${FORENAME}.${LASTNAME}/

* Stefan Jankowski wrote:

> Is this really so? I was under the impression that since the serial
> ports (or at least port A) are mainly meant for plugging terminals
> into, they are actually configured as DCE-ends. 

No, they should be DTEs. DCEs are modems and other things that sit
between endpoints, DTEs are terminating equipment - either a computer
or a terminal or something like that.

I think DTEs were meant to be male and DCEs female, and obviously not
wired the same either.

--tim

On Fri, 29 Apr 2005 11:31:30 +0200, Stefan Jankowski wrote:

>>>>>> Dave Uhring <daveuhring@yahoo.com> writes:
> 
>> The Ultra 1 serial ports are labeled TTY-A and TTY-B but Sun
>> violated common practice for DTE ports by making them DB-25S
>> connectors rather than DB-25P.
> 
> Is this really so? I was under the impression that since the serial
> ports (or at least port A) are mainly meant for plugging terminals
> into, they are actually configured as DCE-ends.

The Ultra 1 was designed as a workstation.  TTY-A logically then would be
used to connect to a modem (DCE) instead.  Although a DCE connector is
used it is wired as DTE.  Hence the requirement for a null modem to
connect to a terminal or terminal emulating PeeCee or other workstation.

> (I never checked the
> specs though, and may be mistaken; incidentially however, on Ultra-5
> and 10 Models (which are otherwise mainly PeeCees with SPARC CPUs)
> port A ist still DB-25S while B is a DE-9P, and the logical conclusion
> seems to be that one is configured as DCE-end for OBP-access and the
> other as DTE to be compatible with the rest of the PeeCee-world...)

No.  Both ports are DTE.  The DB-25S port may be used for both
asynchronous and synchronous connections but it is DTE.  If that were
wired as DCE no null modem would be required to connect to another Sun
machine, yet the null modem is required.

>>>>> Dave Uhring <daveuhring@yahoo.com> writes:

> On Fri, 29 Apr 2005 11:31:30 +0200, Stefan Jankowski wrote:
[...]
>> (I never checked the specs though, and may be mistaken;

Indeed... %)

> No.  Both ports are DTE.  The DB-25S port may be used for both
> asynchronous and synchronous connections but it is DTE.  If
> that were wired as DCE no null modem would be required to
> connect to another Sun machine, yet the null modem is required.

All too valid a point... Taken.

-- 
Stefan Jankowski
Sysadmin, University of Freiburg, CS Dpt.
s/spambucket//
s/dave\.null/${FORENAME}.${LASTNAME}/

"Dave Uhring" <daveuhring@yahoo.com> wrote in message 
news:pan.2005.04.28.18.35.23.835019@yahoo.com...
> On Thu, 28 Apr 2005 09:02:14 -0700, wraezo wrote:
>
>> Well, I don't really know if I did wreck anything, I suppose.  We never
>> have (and likely never will) used the parallel port.
>
> I don't have the schematics for the system board but parallel ports are
> generally TTL circuits (0 - 5VDC).  The serial ports using RS-232 swing
> their voltages between -12VDC and +12VDC.  Applying -12VDC to the output
> of a TTL circuit would probably destroy it.
>
>> It was more for curiosity sake than anything, that I wanted to get on
>> there.  Sun is a bit of a mystery to me so I wanted to explore.  The
>> machine is fairly unimportant in our environment so it wouldn't be a
>> great loss if it died.
>
> The Ultra 1 serial ports are labeled TTY-A and TTY-B but Sun violated
> common practice for DTE ports by making them DB-25S connectors rather than
> DB-25P.
>

Re: last paragraph above:

You must not go back far enough...
Sun follows the proper original "standard" for port gender. PEECEE's (IBM 
and clones) are the ones that violated the gender standards back in the 
early 80's. They also violated the "standard" for parallel by using the D25 
connector instead of the Centronics "standard" C36 to save board space... 
They used the female D25 for the parallel port and then decided to use male 
connectors for the serial ports. Later, they (again to save board space) 
started to use the D9 connector for the serial ports and used a male, 
instead of the more logical female, to be consistent.

Just FYI.
Ed the cable guy 

On Sat, 30 Apr 2005 08:50:31 -0600, Ed Hall wrote:

> You must not go back far enough...

Only 30 years.

> Sun follows the proper original "standard" for port gender. PEECEE's (IBM 
> and clones) are the ones that violated the gender standards back in the 
> early 80's.

Actually, neither RS-232 nor CCITT V.24 specify the gender of the
connectors.  Considering that the PeeCee predates Sun's incorporation are
you sure who violated what?

> They also violated the "standard" for parallel by using the D25 
> connector instead of the Centronics "standard" C36 to save board
> space...

I did the same thing in 1976, but that was because my Altair 8800 case had
cutouts for DB-25 connectors and none for Centronics 36.

> They used the female D25 for the parallel port and then decided to use male 
> connectors for the serial ports.

A good way of distinguishing between the types of ports, which was the
point I was making to the OP.

> Later, they (again to save board space) 
> started to use the D9 connector for the serial ports and used a male, 
> instead of the more logical female, to be consistent.

Every DCE device I have seen uses "female" connectors.  And if you look
closely at a pair of DB-25 connectors you will see why I enclosed female
in quotes.  Which connector shell fits within the other?

"Dave Uhring" <daveuhring@yahoo.com> wrote in message 
news:pan.2005.04.30.16.55.19.598107@yahoo.com...
> On Sat, 30 Apr 2005 08:50:31 -0600, Ed Hall wrote:
>
>> You must not go back far enough...
>
> Only 30 years.

Ah, a relative Newbie ;-)
>
>> Sun follows the proper original "standard" for port gender. PEECEE's (IBM
>> and clones) are the ones that violated the gender standards back in the
>> early 80's.
>
> Actually, neither RS-232 nor CCITT V.24 specify the gender of the
> connectors.  Considering that the PeeCee predates Sun's incorporation are
> you sure who violated what?

Never claimed that Sun set any standard, only meant that in the earlier days 
(prior to either Sun or PC), serial equipment (DTE or DCE) had female 
connectors in LARGE part, cables had male unless it was a male to female 
extension to an existing cable. And, a lot of the "standard" part of many 
standards is defacto, rather than explicit.
>
>> They also violated the "standard" for parallel by using the D25
>> connector instead of the Centronics "standard" C36 to save board
>> space...
>
> I did the same thing in 1976, but that was because my Altair 8800 case had
> cutouts for DB-25 connectors and none for Centronics 36.

You were a true visionary... But did you use a female connector?
>
>> They used the female D25 for the parallel port and then decided to use 
>> male
>> connectors for the serial ports.
>
> A good way of distinguishing between the types of ports, which was the
> point I was making to the OP.
>
>> Later, they (again to save board space)
>> started to use the D9 connector for the serial ports and used a male,
>> instead of the more logical female, to be consistent.
>
> Every DCE device I have seen uses "female" connectors.  And if you look
> closely at a pair of DB-25 connectors you will see why I enclosed female
> in quotes.  Which connector shell fits within the other?
>
If you look at the contacts it is pretty straight forward (pins and holes), 
but was originally a prolonged topic of debate amongst many engineers, I'm 
sure, since it is a matter of viewpoint or reference (shell or contacts). 
Again, a lot is defacto, ever try to figure out genders of coax connectors, 
especially ones embedded in hybrid connectors? 

On Sun, 01 May 2005 08:37:29 -0600, Ed Hall wrote:

> 
> "Dave Uhring" <daveuhring@yahoo.com> wrote in message 
> news:pan.2005.04.30.16.55.19.598107@yahoo.com...

>> Actually, neither RS-232 nor CCITT V.24 specify the gender of the
>> connectors.  Considering that the PeeCee predates Sun's incorporation are
>> you sure who violated what?
> 
> Never claimed that Sun set any standard, only meant that in the earlier days 
> (prior to either Sun or PC), serial equipment (DTE or DCE) had female 
> connectors in LARGE part, cables had male unless it was a male to female 
> extension to an existing cable. And, a lot of the "standard" part of many 
> standards is defacto, rather than explicit.

Which is why I specifically stated "common practice"; there appears to be
no standard specification.  Use of DB-25S connectors on the equipment does
to some extent reduce exposure to bent pins and warranty repairs.

By the time Sun placed its first product on the market IBM had already
established a common practice, one used on a vast majority of computers.

>> I did the same thing in 1976, but that was because my Altair 8800 case had
>> cutouts for DB-25 connectors and none for Centronics 36.
> 
> You were a true visionary... But did you use a female connector?

Of course, for the lower exposure to bent pins on the computer connector,
if by "female" you mean DB-25S.  But I don't recall what pinout I used and
doubt that it was the same as today's standard.

BTW, if Sun could accept that standard from IBM why not the serial
connector standard?

>> Every DCE device I have seen uses "female" connectors.  And if you look
>> closely at a pair of DB-25 connectors you will see why I enclosed female
>> in quotes.  Which connector shell fits within the other?
>>
> If you look at the contacts it is pretty straight forward (pins and holes), 
> but was originally a prolonged topic of debate amongst many engineers, I'm 
> sure, since it is a matter of viewpoint or reference (shell or contacts). 
> Again, a lot is defacto, ever try to figure out genders of coax connectors, 
> especially ones embedded in hybrid connectors?

Well, when I look at the connector shells I see that the S connector fits
inside of the shell of the P connector.  With military connectors
(Amphenol) the gender terms were never used and should not be used now.

Since the D-Sub series of connectors has the "S" and "P" designators the
use of "male" and "female" is redundant and confusing.  Even more
confounding is the description of a DE-9 connector as a DB-9 connector.

Dave Uhring <daveuhring@yahoo.com> writes:

> On Sat, 30 Apr 2005 08:50:31 -0600, Ed Hall wrote:
> 
> > You must not go back far enough...
> 
> Only 30 years.
> 
> > Sun follows the proper original "standard" for port gender. PEECEE's (IBM 
> > and clones) are the ones that violated the gender standards back in the 
> > early 80's.
> 
> Actually, neither RS-232 nor CCITT V.24 specify the gender of the
> connectors.  Considering that the PeeCee predates Sun's incorporation are
> you sure who violated what?

Are you sure about this?

Although I do not have access to either of these standards, all
documents I've seen that references them states that the DCE (Data
Communications Equipment) uses a female connector and a DTE (Data
Terminal Equipment) use a male connector.

The pinout used for the DB-25 equipped Sun systems makes it a DTE,
even though it uses a female connector.

The only systems I've seen that got it right are old Dec systems,
where both terminals and computers had male DB-25.

Thomas

On Sun, 01 May 2005 20:02:22 +0000, Thomas Tornblom wrote:

> Dave Uhring <daveuhring@yahoo.com> writes:

>> Actually, neither RS-232 nor CCITT V.24 specify the gender of the
>> connectors.  Considering that the PeeCee predates Sun's incorporation are
>> you sure who violated what?
> 
> Are you sure about this?

Not 100%.  I cannot find the original documents.

> Although I do not have access to either of these standards, all
> documents I've seen that references them states that the DCE (Data
> Communications Equipment) uses a female connector and a DTE (Data
> Terminal Equipment) use a male connector.

That agrees with every reference to gender which I have found but more
generally:

http://www.uwsg.iu.edu/usail/peripherals/serial/rs232/

"Originally, there was no standardization of the gender of connectors and
serial devices. Both connectors and serial devices could be either gender.
However, since male connectors are more susceptible to injury than female,
most expensive computing hardware these days has female connectors, while
cables are male on both ends."

> The pinout used for the DB-25 equipped Sun systems makes it a DTE,
> even though it uses a female connector.
> 
> The only systems I've seen that got it right are old Dec systems,
> where both terminals and computers had male DB-25.

The original IBM PC used the same convention.  Even though Sun apparently
did not violate a "standard" they certainly are contrary to the practice
utilized by the vast majority of machines.

Dave Uhring <daveuhring@yahoo.com> writes:

> On Sun, 01 May 2005 20:02:22 +0000, Thomas Tornblom wrote:
> 
> > Dave Uhring <daveuhring@yahoo.com> writes:
> 
> >> Actually, neither RS-232 nor CCITT V.24 specify the gender of the
> >> connectors.  Considering that the PeeCee predates Sun's incorporation are
> >> you sure who violated what?
> > 
> > Are you sure about this?
> 
> Not 100%.  I cannot find the original documents.
> 
> > Although I do not have access to either of these standards, all
> > documents I've seen that references them states that the DCE (Data
> > Communications Equipment) uses a female connector and a DTE (Data
> > Terminal Equipment) use a male connector.
> 
> That agrees with every reference to gender which I have found but more
> generally:
> 
> http://www.uwsg.iu.edu/usail/peripherals/serial/rs232/

It seems that the original spec did not specify the gender, but later
ones do, and at least EIA232 does:

http://www.camiresearch.com/Data_Com_Basics/RS232_standard.html

> 
> "Originally, there was no standardization of the gender of connectors and
> serial devices. Both connectors and serial devices could be either gender.
> However, since male connectors are more susceptible to injury than female,
> most expensive computing hardware these days has female connectors, while
> cables are male on both ends."
> 
> > The pinout used for the DB-25 equipped Sun systems makes it a DTE,
> > even though it uses a female connector.
> > 
> > The only systems I've seen that got it right are old Dec systems,
> > where both terminals and computers had male DB-25.
> 
> The original IBM PC used the same convention.  Even though Sun apparently
> did not violate a "standard" they certainly are contrary to the practice
> utilized by the vast majority of machines.
> 

Personally I believe that the "vast majority" got it wrong, at least
if we're including ascii terminals. The majority of the ones I've seen
had female connectors, but were normally DTE:s.

Thomas

On Mon, 02 May 2005 08:12:11 +0000, Thomas Tornblom wrote:

> It seems that the original spec did not specify the gender, but later
> ones do, and at least EIA232 does:
> 
> http://www.camiresearch.com/Data_Com_Basics/RS232_standard.html

I -knew- I had it right the first time :-)  Thanks!

> Personally I believe that the "vast majority" got it wrong, at least
> if we're including ascii terminals. The majority of the ones I've seen
> had female connectors, but were normally DTE:s.

Indeed, but my old Wyse 50 had both since it had DTE and DCE connectors.

* Dave Uhring wrote:

> The original IBM PC used the same convention.  Even though Sun apparently
> did not violate a "standard" they certainly are contrary to the practice
> utilized by the vast majority of machines.

However there's a good argument that being contrary to the common
practice is a sensible decision, because the common practice puts a
fragile connector on an expensive bit of equipment, and a robust
connector on a bit of wire.  I'd much rather have the equipment have
the robust connector and pay for a gender changer if I needed to -
it's *extremely* frustrating (& often expensive) to damage a connector
on a machine!

--tim